1. Field of the Invention
The present invention relates to a method of forming a resist pattern used for a photolithography process in the manufacture of a semiconductor device and, more particularly to a method of forming a resist pattern enabling formation of a fine pattern realizing a resolution higher than the resolution limit determined by the exposure wavelength.
2. Description of the Related Art
In recent years, the semiconductor industry has developed rapidly--with the degree of integration, in semiconductor devices quadrupling every three years. Higher integration of semiconductor devices has been achieved by reducing the minimum line width of the circuit constructing an LSI by 70 percent every generation, that is, from 0.5 .mu.m to 0.35 .mu.m and to 0.25 .mu.m. However, as a result of this miniaturization of semiconductor devices, it has become difficult to maintain a high resolution in processes using photolithography.
Generally, the resolution in photolithography is determined by the equation below in which .lambda. represents the wavelength of a light source and NA represents the numerical aperture, EQU Resolution=.alpha..times.(.lambda./NA) (1)
(where, .alpha. represents the process coefficient determined by the resist and other conditions and having a value of usually about 0.5).
The numerical aperture (NA) of a projection optical system is usually 0.5 to 0.6; thus, the limit of resolution in photolithography is about the wavelength of the light source.
It is possible to use a KrF exciter laser (248 nm in wavelength) for producing 0.25 .mu.m line width generation LSIs. For the next 0.18 .mu.m generation or 0.13 .mu.m generation, even when using an ArF excimer laser (193 nm in wavelength) as a light source, there are problems such as the lack of sufficient study of suitable resist materials. At the present time, no promising photolithographic technology for the shorter wavelength has been developed.
The phase shifting technology has been proposed as a means for solving the problem of the resolution being limited to the wavelength of the light source for exposure. According to the phase shifting technique, light passing through adjoining apertures on a mask are shifted in phase by 180.degree. from each other so as to enable formation of fine patterns with a resolution higher than the value obtained in the above equation (1).
In the conventional phase shifting technology, a microstructure referred to as a "phase shifter" was introduced in a reticle or mask and the phase inverted on the mask. A mask having such a phase shifter is generally difficult to produce. Moreover, there is the problem that no effective methods have been studied for inspecting for defects in the produced mask or repairing defects. Due to this, phase, shifting technology has been slow to be commercialized and has not yet been applied to mass production despite its potential advantages.
The principle of the phase shifting technology, however, is to shift the phases of adjacent light by 180.degree.. If a phase shifter can be provided anywhere in the optical path, it is not necessary to position the phase shifter on the mask.
From this point of view, as lithographic techniques not providing phase shifters on the mask but giving the resist the function of phase shifters, there are the methods of forming patterns disclosed in Japanese Unexamined Patent Publication (Kokai) No. 5-198491 and No. 6-13309.
The conventional methods of forming a pattern disclosed in the above publications will be explained below with reference to FIGS. 1A to 1D. These methods employ a two-layer resist process. First, as shown in FIG. 1A, a lower resist 2 and an upper resist 3 are deposited on a semiconductor substrate 1. The upper resist 3 is coated with a thickness by which the phase difference between light passing through the upper resist 3 and light passing outside the resist 3 becomes 180.degree.. That is, the thickness d.sub.a of the upper resist is determined based on EQU d.sub.a =.lambda./{(2(n-1)} (2)
in which n represents the refractive index of the upper resist.
Next, as shown in FIG. 1A, using a reticle 41 as a mask, first exposure Is performed to form a pattern on the upper resist 3 with about the same resolution as the resolution limit of the optical system. At that time, by using a resist for a KrF laser as the upper resist 3 and using a resist for i-line as the lower resist 2 and by using a KrF laser as a light source, the lower resist 2 is not exposed because it has no sensitivity to the light of the first exposure.
As shown in FIG. 1B, after the first exposure, first development is performed to pattern the upper resist 3.
Next, as shown in FIG. 1C, second exposure is performed over the entire surface at a wavelength to which the lower resist 2 has sensitivity. At this time, the pattern of the upper resist 3 formed by the first exposure functions as a phase shifter. Therefore, the intensity of light becomes zero at the edge of the upper resist pattern due to interference. The principle is the same as that of the "chrome-less" phase shifting technique.
Further, if the entire surface is exposed by light of a wavelength to which the upper resist 3 has sensitivity, the upper resist 3 functioning as a phases shifter becomes completely exposed and can be dissolved in the developing solution. If the upper resist 3 and the lower resist 2 are exposed at the same time, as mentioned above, only the lower resist 2 at the edge of the shifter where the intensity of light becomes zero remains.
According to the above method of forming a pattern, it is possible to form a micropattern of a resolution higher than the resolution limit determined by the wavelength of exposure without using a phase shifting mask (1,1, a mask comprising a phase shifter).
Summarizing the problems to be solved by the invention, there are few resist materials in the commercially available resist materials which are suitable for the above conventional method using two-layer resists.
In the two-layer resist process, a silicon-containing resist is often used as the upper resist. When using an upper resist containing silicon, high content of silicon is required in order to secure a sufficient selectivity between the upper resist and the lower resist when transferring the pattern of the upper resist to the lower resist in the second exposure. As a result, the resolution and other characteristics of the resist are sometimes impaired.
On the other hand, even when not using a silicon-containing resist as the upper resist, there are problems. In the two-layer resist process, the upper resist is formed thinly with the object of improving the transparency and the resolution of the upper resist and with the object of increasing the depth of focus when exposing the lower resist. The resist materials tend to dissolve into each other at the interface between the upper resist and the lower resist. In particular, when the upper resist is made thin, the resolution can fall depending on the combination of resist materials.